Multilayer Profiling of MRD in Patients with Relapsed/Refractory CLL Treated with Venetoclax-Based Regimens in a Real-World Setting

Background Clinical trials have shown that patients with relapsed/refractory chronic lymphocytic leukemia (RR-CLL) treated with venetoclax monotherapy (Ven) or venetoclax-rituximab (VenR) often achieve unmeasurable MRD (uMRD) in peripheral blood (57% of VenR-patients achieved uMRD in peripheral blood at the 12 months landmark by FCM and/or ASO-PCR in the RCT MURANO; Kater AP. et al. JCO 2019 1). uMRD is a strong prognostic factor for progression free survival (PFS) in RR-CLL treated with venetoclax, but the rate of uMRD has not been systematically validated in RR-CLL cohorts treated in the real-world setting. Venetoclax imposes a strong selective pressure on the MRD clone, but the magnitude of the genetic drift of MRD persisting under venetoclax has not been quantified or qualified. Resistance to venetoclax can be mediated by selection of BCL2 mutant clones, but it is unknown if they already emerge in persistent cells of the MRD when the disease is in remission. Methods We analyzed peripheral blood of Swiss patients enrolled in the VeRVe study (NCT03342144), a prospective non-interventional observational study assessing effectiveness, safety, and quality of life in real-world patients with RR-CLL treated with Ven or VenR according to local label. MRD was measured by 8-color multi-parametric flow cytometry (FCM) following the European Research Initiative on CLL (ERIC) guidelines (sensitivity 10 -4), by high throughput sequencing (Ig-HTS, clonoSEQ® Assay) on gDNA from PBMCs (sensitivity 10 -6) and by NGS LyV4.0 CAPP-Seq ctDNA Assay on plasma ctDNA (sensitivity 10 -3) at diagnosis and at 12 months of treatment. To overcome the limitations imposed by the small size of MRD under venetoclax treatment, we leveraged ctDNA for clonal evolution profiling and for the detection of BCL2 mutations. Tumor gDNA was extracted from pre-treatment CLL cells purified by flow cytometry. Germline gDNA was extracted from PB leukocytes collected at the time of best MRD remission. Results In total, 28 patients were recruited. At the time of the analysis, 7 patients went off study before the one-year landmark, 4 patients have not yet reached the one-year landmark, resulting in 17 patients (Ven=5, VenR=12) assessable at the one-year landmark [17 analyzed by FCM, 16 by CAPP-seq (1 ongoing) and 12 by Ig-HTS (5 ongoing; 3 could not be evaluated because they failed Ig-HTS QC)]. At the 12 months landmark, uMRD was achieved in 58% of the assessed cases by FCM, 56% by CAPP-seq and 33% by Ig-HTS. FCM and CAPP-seq concordantly defined the MRD status in 76% of cases with κ=0.49 (12% had measurable ctDNA despite uMRD by FCM and 12% lacked ctDNA despite having measurable MRD by FCM). FCM and Ig-HTS concordantly defined the MRD status in 67% of cases with κ=0.40 (33% had measurable MRD by Ig-HTS despite uMRD by FCM). All patients, who had measurable MRD by FCM and/or CAPP-seq, had measurable MRD by Ig-HTS, which instead detected measurable MRD in 11% of patients who scored negative by both FCM and CAPP-seq (κ=0.77). By CAPP-seq of gDNA from purified CLL cells collected before therapy, the most frequently mutated genes were TP53, IGLL5, BRAF, SF3B1, NXF1, and ZMYM3. At baseline the concordance in mutation recovery by CAPP-seq of gDNA from purified CLL and ctDNA from plasma was high (73%). Before therapy, BCL2 mutations were never detected in either gDNA from purified CLL cells, or in ctDNA. To overcome the limitations imposed to genomic profiling by the small size of MRD under Ven or VenR treatment, we leveraged ctDNA for tracking the acquisition of BCL2 mutations. A total of 5 longitudinal, post-treatment timepoints spanning up to two years of therapy were analyzed. No patients who had measurable ctDNA at the one-year landmark (n=16) acquired BCL2 mutations. Conclusions The rate of uMRD achievable by venetoclax-based therapy in RR-CLL as measured in this real-world prospective study is congruent with the rate observed in the randomized clinical study MURANO. Ig-HTS has the best diagnostic performance for MRD detection. BCL2 mutations are not detectable in ctDNA after one year of venetoclax-based treatment, confirming that time limited therapy has low risk of selecting resistance. 1.) Kater AP. et al. Fixed Duration of Venetoclax-Rituximab in Relapsed/Refractory Chronic Lymphocytic Leukemia Eradicates Minimal Residual Disease and Prolongs Survival: Post-Treatment Follow-Up of the MURANO Phase III Study. JCO 2019, 7(4):269-277